Assistance with 15V 3A SMPS Circuit Design

[zaim]

Hello EEVblog Electronics Community!

I am seeking assistance with a 15V 3A SMPS circuit design that I recently created. Specifically, I am experiencing issues with the dummy load (75ohm 5W) getting very hot, although the circuit works perfectly with the dummy load. I have attached the circuit design in the PDF file for your reference.

I have noticed that several SMPS designs do not include a dummy load, yet their output voltage remains stable. I would greatly appreciate any guidance or support you can provide in modifying my circuit to eliminate the need for a dummy load and ensure a stable output voltage every time.

Thank you in advance for your time and assistance in this matter.

Best regards,
Zaim

[tooki]

I can't advise on how to improve the circuit, but I'm a bit puzzled by your statement:

Specifically, I am experiencing issues with the dummy load (75ohm 5W) getting very hot, although the circuit works perfectly with the dummy load.

By "works perfectly" I assume you mean that you are getting 15V, and that it can handle a 3A load.

Assuming this, why are you surprised that the resistor gets hot?
15V/75 ohms = 0.2A
0.2A * 15V = 3W

3W is plenty of power to heat up a resistor. So I'm curious as to what you think should be happening and why.

[moffy]

Have you tried it with a higher value dummy load, e.g. 150R, if so how was the regulation? A dummy load on a flyback is to prevent a ringing peak being rectified and held as the output on low loads, low ringing the lower power needed for the dummy load.

[zaim]

Dear tooki,

Thank you for your comment and for pointing out the power calculation for the dummy load. You are correct that the 5W 75ohm resistor should be dissipating approximately 3W of power and therefore getting hot.

However, my concern is not with the power dissipation itself, but with the fact that the resistor is getting much hotter than expected. In my experience with other similar circuits, the dummy load typically only gets slightly warm to the touch, whereas in my circuit it is becoming too hot to touch.

So my question is not why the dummy load is dissipating 3W of power, but why it is getting so much hotter than expected with that power dissipation. I suspect there may be an issue with the circuit design or component selection that is causing this excessive heating, and I am hoping for some guidance in identifying and addressing the problem.

Thank you again for your input and any further insight you may have on this matter.

[zaim]

Dear moffy,

I have indeed tried different values of dummy load resistors, including a 150ohm resistor as you suggested.However, I have found that lower values of dummy load resistors, such as the 75ohm resistor I am currently using, provide better and more stable regulation of the output voltage. When using a higher value dummy load, the output voltage tends to fluctuate more and is less consistent under varying loads.

I appreciate your input and any further advice you may have on improving the stability of the circuit without the need for a dummy load. Thank you!

[inse]

If your power supply delivers 15V then it’s neither an issue with the voltage nor the resistance, it’s an issue of your component selection.
Is the resistor the type you screw onto a panel, then do so and make sure it’s metal.
Is it the rectangular ceramic type, then those are allowed to get very hot.

[tooki]

Dear tooki,

Thank you for your comment and for pointing out the power calculation for the dummy load. You are correct that the 5W 75ohm resistor should be dissipating approximately 3W of power and therefore getting hot.

However, my concern is not with the power dissipation itself, but with the fact that the resistor is getting much hotter than expected. In my experience with other similar circuits, the dummy load typically only gets slightly warm to the touch, whereas in my circuit it is becoming too hot to touch.

So my question is not why the dummy load is dissipating 3W of power, but why it is getting so much hotter than expected with that power dissipation. I suspect there may be an issue with the circuit design or component selection that is causing this excessive heating, and I am hoping for some guidance in identifying and addressing the problem.

Thank you again for your input and any further insight you may have on this matter.

I think there's still a disconnect in your logic. You seem to think that 3W dissipation should have the resistor stay cool, and that the heat must be due to some other problem, thus causing higher dissipation. But you said that the circuit is working fine, so I assume that means the output is in fact 15V. (Please confirm.)

If the output is in fact 15V, and the resistor is in fact 75 ohms, then the dissipation is 3W -- it can't be anything else.

So the real question is: why do you think 3W of dissipation shouldn't get hot? Into a small resistor, that is a LOT of power. You haven't specified what kind of resistor it is, but many power resistors will have a temperature rise graph in the datasheet. For example, for the Vishay AC05 axial 5W resistor, at 3W, a temperature rise of about 180K is expected. So it'll be damned hot. A Yageo SPQ500 5W cement resistor (those little white blocks) "only" has a temperature rise of about 120K at 3W. (Why the difference? The cement resistor has a much larger surface area from which to dissipate heat.)

But you see that in either case, the amount of heating that is expected is enough to make it too hot to touch. (Assuming 20C ambient, you're looking at the resistor reaching 140-200 degrees C!)

So it sounds to me like the problem here is that you don't have any feel for how much heat a particular amount of power dissipation should generate, so your "expected" temperature is completely unrealistic for the power dissipation you selected.

If the other designs you're comparing to have the dummy load resistors staying cool, then either:
- they're using physically larger resistors that can dissipate a lot more heat, or
- they're using heatsinks or cooling that can help dissipate more heat, or
- they're simply dissipating a lot less power to begin with

Look at those other designs and see what resistor value they used, and then run the math to see how much power they're dissipating. I suspect you'll find they're dissipating a lot less power, either by using a much higher resistor value, or a much smaller output voltage. (For example, at 5V, a 75 ohm resistor dissipates just 1/3 of a watt, which would only produce about 25K of temperature rise in the cement block resistor above.)

[moffy]

My previous reply might have been a bit simplistic, at low loads you also have a low pulse width which will have a certain amount of jitter because of noise, especially if the di/dt, across the main winding and Q1, is high. Your control loop can become a little unstable because it is effectively seeing a higher loop gain at light loads. You could try reducing the feedback loop gain to see if it helps stabilise the output at light loads, but this would also increase the regulation or output droop of the supply, not sure how much or what is acceptable.